Quantitative assessment of single-cell RNA-sequencing methods

A computational method to aid the design and analysis of single cell RNA-seq experiments for cell type identification

Background

The advent of single cell RNA sequencing (scRNA-seq) enabled researchers to study transcriptomic activity within individual cells and identify inherent cell types in the sample. Although numerous computational tools have been developed to analyze single cell transcriptomes, there are no published studies and analytical packages available to guide experimental design and to devise suitable analysis procedure for cell type identification.

Results

We have developed an empirical methodology to address this important gap in single cell experimental design and analysis into an easy-to-use tool called SCEED (Single Cell Empirical Experimental Design and analysis). With SCEED, user can choose a variety of combinations of tools for analysis, conduct performance analysis of analytical procedures and choose the best procedure, and estimate sample size (number of cells to be profiled) required for a given analytical procedure at varying levels of cell type rarity and other experimental parameters. Using SCEED, we examined 3 single cell algorithms using 48 simulated single cell datasets that were generated for varying number of cell types and their proportions, number of genes expressed per cell, number of marker genes and their fold change, and number of single cells successfully profiled in the experiment.

Conclusions

Based on our study, we found that when marker genes are expressed at fold change of 4 or more, either Seurat or SIMLR algorithm can be used to analyze single cell dataset for any number of single cells isolated (minimum 1000 single cells were tested). However, when marker genes are expected to be only up to fold change of 2, choice of the single cell algorithm is dependent on the number of single cells isolated and rarity of cell types to be identified. In conclusion, our work allows the assessment of various single cell methods and also aids in the design of single cell experiments.

Electronic supplementary material

The online version of this article (10.1186/s12859-019-2817-2) contains supplementary material, which is available to authorized users.

A guide to visualizing the spatial epigenome with super-resolution microscopy

Genomic DNA in eukaryotic cells is tightly compacted with histone proteins into nucleosomes, which are further packaged into the higher-order chromatin structure. The physical structuring of chromatin is highly dynamic and regulated by a large number of epigenetic modifications in response to various environmental exposures, both in normal development and pathological processes such as aging and cancer. Higher-order chromatin structure has been indirectly inferred by conventional bulk biochemical assays on cell populations, which do not allow direct visualization of the spatial information of epigenomics (referred to as spatial epigenomics). With recent advances in super-resolution microscopy, the higher-order chromatin structure can now be visualized in vivo at an unprecedent resolution. This opens up new opportunities to study physical compaction of 3D chromatin structure in single cells, maintaining a well-preserved spatial context of tissue microenvironment. This review discusses the recent application of super-resolution fluorescence microscopy to investigate the higher-order chromatin structure of different epigenomic states. We also envision the synergistic integration of super-resolution microscopy and high-throughput genomic technologies for the analysis of spatial epigenomics to fully understand the genome function in normal biological processes and diseases.

A Microfluidic System for One-Chip Harvesting of Single-Cell-Laden Hydrogels in Culture Medium

Single-cell analysis has shown great potential to fully quantify the distribution of cellular behaviors among a population of individuals. Through isolation and preservation of single cells in the aqueous phase, droplet encapsulation followed by gelation enables high-throughput analysis in biocompatible microgels. However, the lack of control over the number of cells encapsulated and complicated gelation processes significantly limit its efficiency. Here, a microfluidic system for one-chip harvesting of single-cell-laden microgels is presented. Through ultraviolet irradiation, an on-chip gelation technique is seamlessly combined with droplet generation to realize high-throughput fabrication of microscale hydrogels in microfluidic channel. Moreover, a sorting module is introduced to simultaneously complete cell-laden microgel selection and transfer into culture medium. To demonstrate the efficiency of this method, two types of single cells are respectively encapsulated and collected, showing desirable single-cell encapsulation and cell viability. This technique realizes integrated droplet gelation, microgel sorting, and transfer into culture medium, allowing high-throughput analysis of single cells and comprehensive understanding of the cellular specificity.

Differential expression of heat shock and floral regulatory genes in pseudocarpel initials of mantled female inflorescences from Elaeis guineensis Jacq

Transcriptomes generated by laser capture microdissected abnormal staminodes revealed adoption of carpel programming during organ initiation with decreased expression of numerousHSPs,EgDEF1, EgGLO1but increasedLEAFYexpression. The abnormal mantled phenotype in oil palm involves a feminization of the male staminodes into pseudocarpels in pistillate inflorescences. Previous studies on oil palm flowering utilized entire inflorescences or spikelets, which comprised not only the male and female floral organs, but the surrounding tissues as well. Laser capture microdissection coupled with RNA sequencing was conducted to investigate the specific transcriptomes of male and female floral organs from normal and mantled female inflorescences. A higher number of differentially expressed genes (DEGs) were identified in abnormal versus normal male organs compared with abnormal versus normal female organs. In addition, the abnormal male organ transcriptome closely mimics the transcriptome of abnormal female organ. While the transcriptome of abnormal female organ was relatively similar to the normal female organ, a substantial amount of female DEGs encode HEAT SHOCK PROTEIN genes (HSPs). A similar high amount (20%) of male DEGs encode HSPs as well. As these genes exhibited decreased expression in abnormal floral organs, mantled floral organ development may be associated with lower stress indicators. Stamen identity genes EgDEF1 and EgGLO1 were the main floral regulatory genes with decreased expression in abnormal male organs or pseudocarpel initials. Expression of several floral transcription factors was elevated in pseudocarpel initials, notably LEAFY, FIL and DL orthologs, substantiating the carpel specification programming of abnormal staminodes. Specific transcriptomes thus obtained through this approach revealed a host of differentially regulated genes in pseudocarpel initials compared to normal male staminodes.

Dean Flow Assisted Single Cell and Bead Encapsulation for High Performance Single Cell Expression Profiling

Droplet microfluidics-based platform (Drop-seq) has been shown to be a powerful tool for single cell expression profiling. Nevertheless, this platform required the simultaneous encapsulation of single cell and single barcoded bead, the incidence of which was very low, limiting its efficiency. Spiral channels were reported to focus the barcoded beads and thus increased the efficiency, but focusing of cells was not demonstrated, which could potentially further enhance the performance. Here, we designed spiral and serpentine channels to focus both bead and cell solutions and implemented this microfluidic design on Drop-seq. We characterized the effect of cell/bead concentration on encapsulation results and tested the performance by coencapsulating barcoded beads and human-mouse cell mixtures followed by sequencing. The results showed ∼300% and ∼40% increase in cell utilization rate compared to the traditional Drop-seq device and the device focusing beads alone, respectively. This chip design showed great potential for high efficiency single cell expression profiling.

MARS-seq2.0: an experimental and analytical pipeline for indexed sorting combined with single-cell RNA sequencing

Human tissues comprise trillions of cells that populate a complex space of molecular phenotypes and functions and that vary in abundance by 4-9 orders of magnitude. Relying solely on unbiased sampling to characterize cellular niches becomes infeasible, as the marginal utility of collecting more cells diminishes quickly. Furthermore, in many clinical samples, the relevant cell types are scarce and efficient processing is critical. We developed an integrated pipeline for index sorting and massively parallel single-cell RNA sequencing (MARS-seq2.0) that builds on our previously published MARS-seq approach. MARS-seq2.0 is based on >1 million cells sequenced with this pipeline and allows identification of unique cell types across different tissues and diseases, as well as unique model systems and organisms. Here, we present a detailed step-by-step procedure for applying the method. In the improved procedure, we combine sub-microliter reaction volumes, optimization of enzymatic mixtures and an enhanced analytical pipeline to substantially lower the cost, improve reproducibility and reduce well-to-well contamination. Data analysis combines multiple layers of quality assessment and error detection and correction, graphically presenting key statistics for library complexity, noise distribution and sequencing saturation. Importantly, our combined FACS and single-cell RNA sequencing (scRNA-seq) workflow enables intuitive approaches for depletion or enrichment of cell populations in a data-driven manner that is essential to efficient sampling of complex tissues. The experimental protocol, from cell sorting to a ready-to-sequence library, takes 2-3 d. Sequencing and processing the data through the analytical pipeline take another 1-2 d.

Deep transcriptome analysis of the heat shock response in an Atlantic sturgeon (Acipenser oxyrinchus) cell line

Despite efforts to restore Atlantic sturgeon in European rivers, aquaculture techniques result in animals with high post-release mortality due to, among other reasons, their low tolerance to increasing water temperature. Marker genes to monitor heat stress are needed in order to identify heat-resistant fish. Therefore, an Atlantic sturgeon cell line was exposed to different heat shock protocols (30 °C and 35 °C) and differences in gene expression were investigated. In total 3020 contigs (∼1.5%) were differentially expressed. As the core of the upregulated contigs corresponded to heat shock proteins (HSP), the heat shock factor (HSF) and the HSP gene families were annotated in Atlantic sturgeon and mapped via Illumina RNA sequencing to identify heat-inducible family members. Up to 6 hsf and 76 hsp genes were identified in the Atlantic sturgeon transcriptome resources, 16 of which were significantly responsive to the applied heat shock. The previously studied hspa1 (hsp70) gene was only significantly upregulated at the highest heat shock (35 °C), while a set of 5 genes (hspc1, hsph3a, hspb1b, hspb11a, and hspb11b) was upregulated at all conditions. Although the hspc1 (hsp90a) gene was previously used as heat shock-marker in sturgeons, we found that hspb11a is the most heat-inducible gene, with up to 3296-fold higher expression in the treated cells, constituting the candidate gene markers for in vivo trials.

A niche-dependent myeloid transcriptome signature defines dormant myeloma cells

The era of targeted therapies has seen significant improvements in depth of response, progression-free survival, and overall survival for patients with multiple myeloma. Despite these improvements in clinical outcome, patients inevitably relapse and require further treatment. Drug-resistant dormant myeloma cells that reside in specific niches within the skeleton are considered a basis of disease relapse but remain elusive and difficult to study. Here, we developed a method to sequence the transcriptome of individual dormant myeloma cells from the bones of tumor-bearing mice. Our analyses show that dormant myeloma cells express a distinct transcriptome signature enriched for immune genes and, unexpectedly, genes associated with myeloid cell differentiation. These genes were switched on by coculture with osteoblastic cells. Targeting AXL, a gene highly expressed by dormant cells, using small-molecule inhibitors released cells from dormancy and promoted their proliferation. Analysis of the expression of AXL and coregulated genes in human cohorts showed that healthy human controls and patients with monoclonal gammopathy of uncertain significance expressed higher levels of the dormancy signature genes than patients with multiple myeloma. Furthermore, in patients with multiple myeloma, the expression of this myeloid transcriptome signature translated into a twofold increase in overall survival, indicating that this dormancy signature may be a marker of disease progression. Thus, engagement of myeloma cells with the osteoblastic niche induces expression of a suite of myeloid genes that predicts disease progression and that comprises potential drug targets to eradicate dormant myeloma cells.

Unmasking cellular response of a bloom-forming alga to viral infection by resolving expression profiles at a single-cell level

Infection by large dsDNA viruses can lead to a profound alteration of host transcriptome and metabolome in order to provide essential building blocks to support the high metabolic demand for viral assembly and egress. Host response to viral infection can typically lead to diverse phenotypic outcome that include shift in host life cycle and activation of anti-viral defense response. Nevertheless, there is a major bottleneck to discern between viral hijacking strategies and host defense responses when averaging bulk population response. Here we study the interaction between Emiliania huxleyi, a bloom-forming alga, and its specific virus (EhV), an ecologically important host-virus model system in the ocean. We quantified host and virus gene expression on a single-cell resolution during the course of infection, using automatic microfluidic setup that captures individual algal cells and multiplex quantitate PCR. We revealed high heterogeneity in viral gene expression among individual cells. Simultaneous measurements of expression profiles of host and virus genes at a single-cell level allowed mapping of infected cells into newly defined infection states and allowed detection specific host response in a subpopulation of infected cell which otherwise masked by the majority of the infected population. Intriguingly, resistant cells emerged during viral infection, showed unique expression profiles of metabolic genes which can provide the basis for discerning between viral resistant and susceptible cells within heterogeneous populations in the marine environment. We propose that resolving host-virus arms race at a single-cell level will provide important mechanistic insights into viral life cycles and will uncover host defense strategies.

Almost all of our current understanding of the molecular mechanisms that govern host-pathogen interactions in the ocean is derived from experiments carried out at the population level, neglecting any heterogeneity. Here we used a single cell approach to unmask the phenotypic heterogeneity produced within infected populations of the cosmopolitan bloom-forming alga Emiliania huxleyi by its specific lytic virus. We found high variability in expression of viral genes among individual cells. This heterogeneity was used to map cells into their infection state and allowed to uncover a yet unrecognized host response. We also provide evidence that variability in host metabolic states provided a sensitive tool to decipher between susceptible and resistant cells.

Opto-magnetic capture of individual cells based on visual phenotypes

The ability to isolate rare live cells within a heterogeneous population based solely on visual criteria remains technically challenging, due largely to limitations imposed by existing sorting technologies. Here, we present a new method that permits labeling cells of interest by attaching streptavidin-coated magnetic beads to their membranes using the lasers of a confocal microscope. A simple magnet allows highly specific isolation of the labeled cells, which then remain viable and proliferate normally. As proof of principle, we tagged, isolated, and expanded individual cells based on three biologically relevant visual characteristics: i) presence of multiple nuclei, ii) accumulation of lipid vesicles, and iii) ability to resolve ionizing radiation-induced DNA damage foci. Our method constitutes a rapid, efficient, and cost-effective approach for isolation and subsequent characterization of rare cells based on observable traits such as movement, shape, or location, which in turn can generate novel mechanistic insights into important biological processes.

Analysis of Single-Cell RNA-seq Data by Clustering Approaches

Background:

The recently developed single-cell RNA sequencing (scRNA-seq) has attracted a great amount of attention due to its capability to interrogate expression of individual cells, which is superior to traditional bulk cell sequencing that can only measure mean gene expression of a population of cells. scRNA-seq has been successfully applied in finding new cell subtypes. New computational challenges exist in the analysis of scRNA-seq data.

Objective:

We provide an overview of the features of different similarity calculation and clustering methods, in order to facilitate users to select methods that are suitable for their scRNA-seq. We would also like to show that feature selection methods are important to improve clustering performance.

Results:

We first described similarity measurement methods, followed by reviewing some new clustering methods, as well as their algorithmic details. This analysis revealed several new questions, including how to automatically estimate the number of clustering categories, how to discover novel subpopulation, and how to search for new marker genes by using feature selection methods.

Conclusion:

Without prior knowledge about the number of cell types, clustering or semisupervised learning methods are important tools for exploratory analysis of scRNA-seq data.</P>

Comparative analysis of sequencing technologies for single-cell transcriptomics

Single-cell RNA-seq technologies require library preparation prior to sequencing. Here, we present the first report to compare the cheaper BGISEQ-500 platform to the Illumina HiSeq platform for scRNA-seq. We generate a resource of 468 single cells and 1297 matched single cDNA samples, performing SMARTer and Smart-seq2 protocols on two cell lines with RNA spike-ins. We sequence these libraries on both platforms using single- and paired-end reads. The platforms have comparable sensitivity and accuracy in terms of quantification of gene expression, and low technical variability. Our study provides a standardized scRNA-seq resource to benchmark new scRNA-seq library preparation protocols and sequencing platforms.

A fast and efficient count-based matrix factorization method for detecting cell types from single-cell RNAseq data

Background

Single-cell RNA sequencing (scRNAseq) data always involves various unwanted variables, which would be able to mask the true signal to identify cell-types. More efficient way of dealing with this issue is to extract low dimension information from high dimensional gene expression data to represent cell-type structure. In the past two years, several powerful matrix factorization tools were developed for scRNAseq data, such as NMF, ZIFA, pCMF and ZINB-WaVE. But the existing approaches either are unable to directly model the raw count of scRNAseq data or are really time-consuming when handling a large number of cells (e.g. n>500).

Results

In this paper, we developed a fast and efficient count-based matrix factorization method (single-cell negative binomial matrix factorization, scNBMF) based on the TensorFlow framework to infer the low dimensional structure of cell types. To make our method scalable, we conducted a series of experiments on three public scRNAseq data sets, brain, embryonic stem, and pancreatic islet. The experimental results show that scNBMF is more powerful to detect cell types and 10 - 100 folds faster than the scRNAseq bespoke tools.

Conclusions

In this paper, we proposed a fast and efficient count-based matrix factorization method, scNBMF, which is more powerful for detecting cell type purposes. A series of experiments were performed on three public scRNAseq data sets. The results show that scNBMF is a more powerful tool in large-scale scRNAseq data analysis. scNBMF was implemented in R and Python, and the source code are freely available at https://github.com/sqsun.

A diffusion-based microfluidic device for single-cell RNA-seq

Microfluidic devices provide a low-input and efficient platform for single-cell RNA-seq (scRNA-seq). Existing microfluidic devices have a complicated multi-chambered structure for handling the multi-step process involved in RNA-seq and dilution between steps is used to negate the inhibitory effects among reagents. This makes the device difficult to fabricate and operate. Here we present microfluidic diffusion-based RNA-seq (MID-RNA-seq) for conducting scRNA-seq with a diffusion-based reagent swapping scheme. This device incorporates cell trapping, lysis, reverse transcription and PCR amplification all in one simple microfluidic device. MID-RNA-seq provides high data quality that is comparable to existing scRNA-seq methods while implementing a simple device design that permits multiplexing. The robustness and scalability of the MID-RNA-seq device will be important for transcriptomic studies of scarce cell samples.

Transcriptomic behavior of genes associated with chromosome 21 aneuploidies in early embryo development

OBJECTIVE

To analyze how chromosome 21 (HSA21) ploidy affects global gene expression of early human blastocysts.

DESIGN

Prospective study.

SETTING

University-affiliated in vitro fertilization clinic.

PATIENT(S)

A total of 26 high-quality donated embryos from in vitro fertilization (IVF) patients: trisomy 21 (n = 8), monosomy 21 (n = 10), and euploid (n = 8) blastocysts.

INTERVENTION(S)

None.

MAIN OUTCOME MEASURE(S)

Blastocyst transcriptome changes and its associated functions.

RESULT(S)

Trisomy 21, monosomy 21, and euploid blastocysts were classified by comparative genomic hybridization. The global transcriptome of whole blastocysts was analyzed with small cell number RNA sequencing, and they were compared to understand the gene expression behavior at early development and its implications for embryo implantation. We identified 1,232 differentially expressed genes (false discovery rate <0.05) in monosomy 21 compared with euploid blastocysts associated with dysregulated functions in embryo development as the Rap1 signaling pathway. Curiously, Down syndrome in early development revealed fewer transcriptomic changes than expected. In addition, Down syndrome gene expression in neonates, children, and adults revealed that the number of deregulated genes increases across life stages from blastocysts to adults, suggesting a potential dosage-compensation mechanism for human chromosome 21.

CONCLUSION(S)

At the transcriptomic level, early development in Down syndrome is mainly dosage compensated. However, monosomy 21 is strongly transcriptionally affected because early development involving main functions is associated with embryo implantation.

Single-Cell Omics Analyses Enabled by Microchip Technologies

Single-cell omics studies provide unique information regarding cellular heterogeneity at various levels of the molecular biology central dogma. This knowledge facilitates a deeper understanding of how underlying molecular and architectural changes alter cell behavior, development, and disease processes. The emerging microchip-based tools for single-cell omics analysis are enabling the evaluation of cellular omics with high throughput, improved sensitivity, and reduced cost. We review state-of-the-art microchip platforms for profiling genomics, epigenomics, transcriptomics, proteomics, metabolomics, and multi-omics at single-cell resolution. We also discuss the background of and challenges in the analysis of each molecular layer and integration of multiple levels of omics data, as well as how microchip-based methodologies benefit these fields. Additionally, we examine the advantages and limitations of these approaches. Looking forward, we describe additional challenges and future opportunities that will facilitate the improvement and broad adoption of single-cell omics in life science and medicine.

A brief review of single-cell transcriptomic technologies

In recent years, there has been an effort to develop new technologies for measuring gene expression and sequence information from thousands of individual cells. Large data sets that were obtained using these 'single cell' technologies have allowed scientists to address fundamental questions in biomedicine ranging from stems cells and development to cancer and immunology. Here, we provide a brief review of recent developments in single-cell technology. Our intention is to provide a quick background for newcomers to the field as well as a deeper description of some of the leading technologies to date.

Single-Cell RNA Sequencing to Understand Host-Pathogen Interactions

Host-pathogen interactions, particularly in the context of bacterial infections, are dynamic exchanges where transcriptional heterogeneity from both the host and the pathogen can lead to many diverse outcomes via distinct molecular pathways. Transcriptional profiling at the single-cell level, on a genome-wide scale, has enabled a greater appreciation of the cellular diversity in complex biological organisms and the myriad of host transcriptional states during infection. Here, we highlight recent reports of single-cell RNA sequencing within the context of host-pathogen interactions, describe current limitations for detecting and profiling the transcriptome of invading pathogens at the single-cell level, and suggest exciting future prospects for this technology in the study of infection. We propose that understanding infection as an integrated process between pathogen and host with resolution at the single-cell level will ultimately inform development of vaccines with greater productive and protective host immunity, enable the development of novel therapeutics that harness host mechanisms, and yield more accurate biomarkers to guide better diagnostics.

West Nile Virus-Inclusive Single-Cell RNA Sequencing Reveals Heterogeneity in the Type I Interferon Response within Single Cells

West Nile virus (WNV) is a neurotropic mosquito-borne flavivirus of global importance. Neuroinvasive WNV infection results in encephalitis and can lead to prolonged neurological impairment or death. Type I interferon (IFN-I) is crucial for promoting antiviral defenses through the induction of antiviral effectors, which function to restrict viral replication and spread. However, our understanding of the antiviral response to WNV infection is mostly derived from analysis of bulk cell populations. It is becoming increasingly apparent that substantial heterogeneity in cellular processes exists among individual cells, even within a seemingly homogenous cell population. Here, we present WNV-inclusive single-cell RNA sequencing (scRNA-seq), an approach to examine the transcriptional variation and viral RNA burden across single cells. We observed that only a few cells within the bulk population displayed robust transcription of IFN-β mRNA, and this did not appear to depend on viral RNA abundance within the same cell. Furthermore, we observed considerable transcriptional heterogeneity in the IFN-I response, with genes displaying high unimodal and bimodal expression patterns. Broadly, IFN-stimulated genes negatively correlated with viral RNA abundance, corresponding with a precipitous decline in expression in cells with high viral RNA levels. Altogether, we demonstrated the feasibility and utility of WNV-inclusive scRNA-seq as a high-throughput technique for single-cell transcriptomics and WNV RNA detection. This approach can be implemented in other models to provide insights into the cellular features of protective immunity and identify novel therapeutic targets.IMPORTANCE West Nile virus (WNV) is a clinically relevant pathogen responsible for recurrent epidemics of neuroinvasive disease. Type I interferon is essential for promoting an antiviral response against WNV infection; however, it is unclear how heterogeneity in the antiviral response at the single-cell level impacts viral control. Specifically, conventional approaches lack the ability to distinguish differences across cells with varying viral abundance. The significance of our research is to demonstrate a new technique for studying WNV infection at the single-cell level. We discovered extensive variation in antiviral gene expression and viral abundance across cells. This protocol can be applied to primary cells or in vivo models to better understand the underlying cellular heterogeneity following WNV infection for the development of targeted therapeutic strategies.

Single-Cell RNA-Seq of the Pancreatic Islets--a Promise Not yet Fulfilled?

In the past 3 years, we have seen a flurry of publications on single-cell RNA sequencing (RNA-seq) analyses of pancreatic islets from mouse and human. This technology holds the promise to refine cell-type signatures and discover cellular heterogeneity among the canonical endocrine cell types such as the glucagon-producing α and insulin-producing β cells, going as far as suggesting new subtypes. In addition, single-cell RNA-seq has the ability to characterize rare endocrine cell types that are not captured by prior bulk analysis. With transcriptomics data from individual endocrine cells, cellular states can be profiled both along developmental processes and during the emergence of metabolic diseases. However, the promises of this new technology have not yet been met in full. While the methodology for the first time enabled the transcriptional definition of rare endocrine cell types such as ghrelin-producing ɛ cells, some of the conclusions regarding cell-type-specific gene expression changes in type 2 diabetes might need to be revisited once larger sample sizes become available. Data generation and analysis are continuously improving single-cell RNA-seq approaches and are helping us to understand the (mal)adaptations of the islet cells during development, metabolic challenge, and disease.

SinNLRR: a robust subspace clustering method for cell type detection by non-negative and low-rank representation

Motivation

The development of single-cell RNA-sequencing (scRNA-seq) provides a new perspective to study biological problems at the single-cell level. One of the key issues in scRNA-seq analysis is to resolve the heterogeneity and diversity of cells, which is to cluster the cells into several groups. However, many existing clustering methods are designed to analyze bulk RNA-seq data, it is urgent to develop the new scRNA-seq clustering methods. Moreover, the high noise in scRNA-seq data also brings a lot of challenges to computational methods.

Results

In this study, we propose a novel scRNA-seq cell type detection method based on similarity learning, called SinNLRR. The method is motivated by the self-expression of the cells with the same group. Specifically, we impose the non-negative and low rank structure on the similarity matrix. We apply alternating direction method of multipliers to solve the optimization problem and propose an adaptive penalty selection method to avoid the sensitivity to the parameters. The learned similarity matrix could be incorporated with spectral clustering, t-distributed stochastic neighbor embedding for visualization and Laplace score for prioritizing gene markers. In contrast to other scRNA-seq clustering methods, our method achieves more robust and accurate results on different datasets.

Availability and implementation

Our MATLAB implementation of SinNLRR is available at, https://github.com/zrq0123/SinNLRR.

Supplementary information

Supplementary data are available at Bioinformatics online.

The impact of estimated tumour purity on gene expression-based drug repositioning of Clear Cell Renal Cell Carcinoma samples

To find new potentially therapeutic drugs against clear cell Renal Cell Carcinoma (ccRCC), within drugs currently prescribed for other diseases (drug repositioning), we previously searched for drugs which are expected to bring the gene expression of 500 + ccRCC samples from The Cancer Genome Atlas closer to that of healthy kidney tissue samples. An inherent limitation of this bulk RNA-seq data is that tumour samples consist of a varying mixture of cancerous and non-cancerous cells, which influences differential gene expression analyses. Here, we investigate whether the drug repositioning candidates are expected to target the genes dysregulated in ccRCC cells by studying the association with tumour purity. When all ccRCC samples are analysed together, the drug repositioning potential of identified drugs start decreasing above 80% estimated tumour purity. Because ccRCC is a highly vascular tumour, attributed to frequent loss of VHL function and subsequent activation of Hypoxia-Inducible Factor (HIF), we stratified the samples by observed activation of the HIF-pathway. After stratification, the association between estimated tumour purity and drug repositioning potential disappears for HIF-activated samples. This result suggests that the identified drug repositioning candidates specifically target the genes expressed by HIF-activated ccRCC tumour cells, instead of genes expressed by other cell types part of the tumour micro-environment.

Analysis of mRNA abundance for histone variants, histone- and DNA-modifiers in bovine in vivo and in vitro oocytes and embryos

Transcript abundance of histone variants, modifiers of histone and DNA in bovine in vivo oocytes and embryos were measured as mean transcripts per million (TPM). Six of 14 annotated histone variants, 8 of 52 histone methyl-transferases, 5 of 29 histone de-methylases, 5 of 20 acetyl-transferases, 5 of 19 de-acetylases, 1 of 4 DNA methyl-transferases and 0 of 3 DNA de-methylases were abundant (TPM >50) in at least one stage studied. Overall, oocytes and embryos contained more varieties of mRNAs for histone modification than for DNA. Three expression patterns were identified for histone modifiers: (1) transcription before embryonic genome activation (EGA) and down-regulated thereafter such as PRMT1; (2) low in oocytes but transiently increased for EGA such as EZH2; (3) high in oocytes but decreased by EGA such as SETD3. These expression patterns were altered by in vitro culture. Additionally, the presence of mRNAs for the TET enzymes throughout pre-implantation development suggests persistent de-methylation. Together, although DNA methylation changes are well-recognized, the first and second orders of significance in epigenetic changes by in vivo embryos may be histone variant replacements and modifications of histones.

Long-Term Perfusion Culture of Monoclonal Embryonic Stem Cells in 3D Hydrogel Beads for Continuous Optical Analysis of Differentiation

Developmental cell biology requires technologies in which the fate of single cells is followed over extended time periods, to monitor and understand the processes of self-renewal, differentiation, and reprogramming. A workflow is presented, in which single cells are encapsulated into droplets (Ø: 80 µm, volume: ≈270 pL) and the droplet compartment is later converted to a hydrogel bead. After on-chip de-emulsification by electrocoalescence, these 3D scaffolds are subsequently arrayed on a chip for long-term perfusion culture to facilitate continuous cell imaging over 68 h. Here, the response of murine embryonic stem cells to different growth media, 2i and N2B27, is studied, showing that the exit from pluripotency can be monitored by fluorescence time-lapse microscopy, by immunostaining and by reverse-transcription and quantitative PCR (RT-qPCR). The defined 3D environment emulates the natural context of cell growth (e.g., in tissue) and enables the study of cell development in various matrices. The large scale of cell cultivation (in 2000 beads in parallel) may reveal infrequent events that remain undetected in lower throughput or ensemble studies. This platform will help to gain qualitative and quantitative mechanistic insight into the role of external factors on cell behavior.

C1 CAGE detects transcription start sites and enhancer activity at single-cell resolution

Single-cell transcriptomic profiling is a powerful tool to explore cellular heterogeneity. However, most of these methods focus on the 3′-end of polyadenylated transcripts and provide only a partial view of the transcriptome. We introduce C1 CAGE, a method for the detection of transcript 5′-ends with an original sample multiplexing strategy in the C1^TM microfluidic system. We first quantifiy the performance of C1 CAGE and find it as accurate and sensitive as other methods in the C1 system. We then use it to profile promoter and enhancer activities in the cellular response to TGF-β of lung cancer cells and discover subpopulations of cells differing in their response. We also describe enhancer RNA dynamics revealing transcriptional bursts in subsets of cells with transcripts arising from either strand in a mutually exclusive manner, validated using single molecule fluorescence in situ hybridization.

Single-cell transcriptomic profiling allows the exploration of cellular heterogeneity but commonly focuses on the 3′-end of the transcript. Here the authors introduce C1 CAGE, which detects the 5′ transcript end in a multiplexed microfluidic system.

Molecular signatures of multiple myeloma progression through single cell RNA-Seq

We used single cell RNA-Seq to examine molecular heterogeneity in multiple myeloma (MM) in 597 CD138 positive cells from bone marrow aspirates of 15 patients at different stages of disease progression. 790 genes were selected by coefficient of variation (CV) method and organized cells into four groups (L1–L4) using unsupervised clustering. Plasma cells from each patient clustered into at least two groups based on gene expression signature. The L1 group contained cells from all MGUS patients having the lowest expression of genes involved in the oxidative phosphorylation, Myc targets, and mTORC1 signaling pathways (p < 1.2 × 10⁻¹⁴). In contrast, the expression level of these pathway genes increased progressively and were the highest in L4 group containing only cells from MM patients with t(4;14) translocations. A 44 genes signature of consistently overexpressed genes among the four groups was associated with poorer overall survival in MM patients (APEX trial, p < 0.0001; HR, 1.83; 95% CI, 1.33–2.52), particularly those treated with bortezomib (p < 0.0001; HR, 2.00; 95% CI, 1.39–2.89). Our study, using single cell RNA-Seq, identified the most significantly affected molecular pathways during MM progression and provided a novel signature predictive of patient prognosis and treatment stratification.

Classification of Single-Cell Gene Expression Trajectories from Incomplete and Noisy Data

This paper studies classification of gene-expression trajectories coming from two classes, healthy and mutated (cancerous) using Boolean networks with perturbation (BNps) to model the dynamics of each class at the state level. Each class has its own BNp, which is partially known based on gene pathways. We employ a Gaussian model at the observation level to show the expression values of the genes given the hidden binary states at each time point. We use expectation maximization (EM) to learn the BNps and the unknown model parameters, derive closed-form updates for the parameters, and propose a learning algorithm. After learning, a plug-in Bayes classifier is used to classify unlabeled trajectories, which can have missing data. Measuring gene expressions at different times yields trajectories only when measurements come from a single cell. In multiple-cell scenarios, the expression values are averages over many cells with possibly different states. Via the central-limit theorem, we propose another model for expression data in multiple-cell scenarios. Simulations demonstrate that single-cell trajectory data can outperform multiple-cell average expression data relative to classification error, especially in high-noise situations. We also consider data generated via a mammalian cell-cycle network, both the wild-type and with a common mutation affecting p27.

Integrated Fluidic Circuits for Single-Cell Omics and Multi-omics Applications

Single-cell genomics plays a crucial role in several aspects of biology, from developmental biology to mapping every cell in the human body through the Cell Atlas initiative. To meet these various applications, single-cell methods are rapidly evolving to increase throughput; improve sensitivity, quantification accuracy, and usability; and reduce nucleic-acid amplification bias and cost. In addition to improvement in single-cell methods, there is a huge interest in analyzing multiple analytes such as genome, epigenome, transcriptome, and protein from the same single cell. This approach is generalized as single-cell multi-omics. Automation of multi-step single-cell methods is highly desired to achieve a reproducible workflow; reduce human error and avoid contamination; and introduce technical variability to an existing stochastic process. Typically single-cell reactions start with a low level of nucleic acid, in the range of picograms. Miniaturization in microfluidic devices leads to a gain in reaction efficiency in Nanoliter or picoliter reaction volumes and active mixing help ensure that solid-state microfluidic devices provide the broadest flexibility and best sensitivity in single-cell reactions, compared to other methods. In this chapter, we will present integrated fluidic circuit (IFC) microfluidics for various single-cell multi-omics applications, and show how this technology fits into the current single-cell technology portfolio available from various vendors. We will then discuss possible uses for IFCs in multi-omics applications that are on the horizon.

Quantitation of mRNA Transcripts and Proteins Using the BD Rhapsody™ Single-Cell Analysis System

In this review, we describe the BD Rhapsody™ Single-Cell Analysis System, a platform that allows high-throughput capture of nucleic acids from single cells using a simple cartridge workflow and a multitier barcoding system. The resulting captured information can be used to generate various types of next-generation sequencing (NGS) libraries, including whole transcriptome analysis for discovery biology and targeted RNA analysis for high sensitivity transcript detection. The BD Rhapsody system can be used with emerging applications, such as BD™ AbSeq assays, to profile gene expression in both mRNA and protein level to provide ultra-high resolution analysis of single cells.

Single-Cell DNA-Seq and RNA-Seq in Cancer Using the C1 System

Heterogeneous phenotypes of cancer cells enable them to adapt to various environments. The heterogeneity results from diversity of genome, transcriptome, and epigenome at a single-cell level. The C1 system can automatically perform single-cell capture and whole genome amplification (WGA) or whole transcription amplification (WTA) by MDA or Smart-Seq, respectively. Here, we describe the protocols for WGA and WTA from a single cell by using the C1 system and the protocols for sequence library preparation from amplified gDNA and cDNA. We also described about the computational analysis for single-cell data of cancer.

Wound Healing: A Cellular Perspective

Wound healing is one of the most complex processes in the human body. It involves the spatial and temporal synchronization of a variety of cell types with distinct roles in the phases of hemostasis, inflammation, growth, re-epithelialization, and remodeling. With the evolution of single cell technologies, it has been possible to uncover phenotypic and functional heterogeneity within several of these cell types. There have also been discoveries of rare, stem cell subsets within the skin, which are unipotent in the uninjured state, but become multipotent following skin injury. Unraveling the roles of each of these cell types and their interactions with each other is important in understanding the mechanisms of normal wound closure. Changes in the microenvironment including alterations in mechanical forces, oxygen levels, chemokines, extracellular matrix and growth factor synthesis directly impact cellular recruitment and activation, leading to impaired states of wound healing. Single cell technologies can be used to decipher these cellular alterations in diseased states such as in chronic wounds and hypertrophic scarring so that effective therapeutic solutions for healing wounds can be developed.

Strategies for Converting RNA to Amplifiable cDNA for Single-Cell RNA Sequencing Methods

This review describes the features of molecular biology techniques for single-cell RNA sequencing (scRNA-seq), including methods developed in our laboratory. Existing scRNA-seq methods require the conversion of first-strand cDNA to amplifiable cDNA followed by whole-transcript amplification. There are three primary strategies for this conversion: poly-A tagging, template switching, and RNase H-DNA polymerase I-mediated second-strand cDNA synthesis for in vitro transcription. We discuss the merits and limitations of these strategies and describe our Reverse Transcription with Random Displacement Amplification technology that allows for direct first-strand cDNA amplification from RNA without the need for conversion to an amplifiable cDNA. We believe that this review provides all users of single-cell transcriptome technologies with an understanding of the relationship between the quantitative performance of various methods and their molecular features.

Varying dietary protein and fat elicits differential transcriptomic expression within stress response pathways in preweaned Holstein heifers

Increases in milk replacer dietary energy subsequently increase growth and weight in preweaned dairy heifers. However, the underlying effects of dietary component increases on key functional pathways have yet to be fully investigated. Elucidating these relationships may provide insights into the mechanisms through which protein and fat are partitioned for tissue growth and metabolism. We hypothesized that genes within key growth and metabolic pathways would be differentially expressed between calves fed a protein- and fat-restricted diet and calves fed a protein- and fat-enhanced diet. The objectives of this study were to (1) identify genes differentially expressed between dietary restricted calves and enhanced calves and (2) determine the key regulatory pathways influenced by these genes. Preweaned Holstein heifers (n = 12; 6 ± 0.02 d of age) were randomly assigned to 1 of 2 milk replacer diets: enhanced (28.9% crude protein, 26.2% fat; n = 6) or restricted (20.9% crude protein, 19.8% fat; n = 6). Growth measures included average daily gain and gain-to-feed ratio. After 56 d, calves were killed for tissue collection. Samples from longissimus dorsi, adipose, and liver tissues were collected and RNA was isolated for RNA sequencing analysis. The MIXED procedure of SAS (SAS Institute Inc., Cary, NC) was used to evaluate relationships of growth with dietary energy. Fixed effects included date of collection and time (day). Random effects included sire and birth weight. The RNA sequencing analysis was performed using CLC Genomics Workbench (Qiagen, Germantown, MD), and the Robinson and Smith exact test was used to identify differentially expressed genes between diets. The Protein Analysis Through Evolutionary Relationships (PANTHER) database was then used to identify functional categories of differentially expressed genes. Enhanced calves had increased growth rates and feed efficiency compared with restricted calves (average daily gain = 0.76 and 0.22, respectively; gain-to-feed ratio = 0.10 and 0.06, respectively). There were 238 differentially expressed genes in adipose, 227 in longissimus dorsi, and 40 in liver. We identified 10 genes concordant among tissues. As expected, functional analyses suggested that the majority of genes were associated with metabolic or cellular processes, predominantly cell communication and cell cycle. Overall, it appears that varying levels of dietary protein and fat influence calf growth and development through metabolic processes, including oxidative phosphorylation and glyceroneogenesis. However, protein- and fat-restricted calves appeared to experience metabolic stress at a cellular level, as evidenced by an upregulation in stress response pathways, including genes in the p53 pathway. Calves could be fed at a higher level of protein and fat to decrease the prevalence of metabolic stress at the cellular level, but evidence indicating the presence of inflammatory stress and adipose fibrosis in enhanced calves prompts further investigation of the effects of milk replacer component levels.

Gene expression effects of lithium and valproic acid in a serotonergic cell line

Valproic acid (VPA) and lithium are widely used in the treatment of bipolar disorder. However, the underlying mechanism of action of these drugs is not clearly understood. We used RNA-Seq analysis to examine the global profile of gene expression in a rat serotonergic cell line (RN46A) after exposure to these two mood stabilizer drugs. Numerous genes were differentially regulated in response to VPA (log₂ fold change ≥ 1.0; i.e., odds ratio of ≥2, at false discovery rate <5%), but only two genes ( Dynlrb2 and Cdyl2) showed significant differential regulation after exposure of the cells to lithium, with the same analysis criteria. Both of these genes were also regulated by VPA. Many of the differentially expressed genes had functions of potential relevance to mood disorders or their treatment, such as several serpin family genes (including neuroserpin), Nts (neurotensin), Maob (monoamine oxidase B), and Ap2b1, which is important for synaptic vesicle function. Pathway analysis revealed significant enrichment of Gene Ontology terms such as extracellular matrix remodeling, cell adhesion, and chemotaxis. This study in a cell line derived from the raphe nucleus has identified a range of genes and pathways that provide novel insights into potential therapeutic actions of the commonly used mood stabilizer drugs.

Applications of Immunomodulatory Immune Synergies to Adjuvant Discovery and Vaccine Development

Pathogens comprise a diverse set of immunostimulatory molecules that activate the innate immune system during infection. The immune system recognizes distinct combinations of pathogenic molecules leading to multiple immune activation events that cooperate to produce enhanced immune responses, known as 'immune synergies'. Effective immune synergies are essential for the clearance of pathogens, thus inspiring novel adjuvant design to improve vaccines. We highlight current vaccine adjuvants and the importance of immune synergies to adjuvant and vaccine design. The focus is on new technologies used to study and apply immune synergies to adjuvant and vaccine development. Finally, we discuss how recent findings can be applied to the future design and characterization of synergistic adjuvants and vaccines.

Integrative analysis of single-cell expression data reveals distinct regulatory states in bidirectional promoters

Background

Bidirectional promoters (BPs) are prevalent in eukaryotic genomes. However, it is poorly understood how the cell integrates different epigenomic information, such as transcription factor (TF) binding and chromatin marks, to drive gene expression at BPs. Single-cell sequencing technologies are revolutionizing the field of genome biology. Therefore, this study focuses on the integration of single-cell RNA-seq data with bulk ChIP-seq and other epigenetics data, for which single-cell technologies are not yet established, in the context of BPs.

Results

We performed integrative analyses of novel human single-cell RNA-seq (scRNA-seq) data with bulk ChIP-seq and other epigenetics data. scRNA-seq data revealed distinct transcription states of BPs that were previously not recognized. We find associations between these transcription states to distinct patterns in structural gene features, DNA accessibility, histone modification, DNA methylation and TF binding profiles.

Conclusions

Our results suggest that a complex interplay of all of these elements is required to achieve BP-specific transcriptional output in this specialized promoter configuration. Further, our study implies that novel statistical methods can be developed to deconvolute masked subpopulations of cells measured with different bulk epigenomic assays using scRNA-seq data.

Electronic supplementary material

The online version of this article (10.1186/s13072-018-0236-7) contains supplementary material, which is available to authorized users.

Simulation-based benchmarking of isoform quantification in single-cell RNA-seq

Single-cell RNA-seq has the potential to facilitate isoform quantification as the confounding factor of a mixed population of cells is eliminated. However, best practice for using existing quantification methods has not been established. We carry out a benchmark for five popular isoform quantification tools. Performance is generally good for simulated data based on SMARTer and SMART-seq2 data. The reduction in performance compared with bulk RNA-seq is small. An important biological insight comes from our analysis of real data which shows that genes that express two isoforms in bulk RNA-seq predominantly express one or neither isoform in individual cells.

Optimised isolation method for RNA extraction suitable for RNA sequencing from feline teeth collected in a clinical setting and at post mortem

Advanced next generation sequencing approaches have started to reveal the cellular and molecular complexity of the microenvironment in many tissues. It is challenging to obtain high quality RNA from mineralised tissues. We developed an optimised method of RNA extraction from feline teeth collected in a clinical setting and at post mortem. Teeth were homogenised in phenol-guanidinium solution at near-freezing temperatures and followed by solid-phase nucleic acid extraction utilising a commercially available kit. This method produced good RNA yields and improved RNA quality based on RNA integrity numbers equivalent (RIN^e) from an average of 3.6 to 5.6. No correlation was found between RNA purity parameters measured by A_260:280 or A_230:260 ratios and degree of RNA degradation. This implies that RNA purity indicators cannot be reliably used as parameters of RNA integrity. Two reference genes (GAPDH, RPS19) showed significant changes in expression levels by qPCR at low and moderate RIN^e values, while RPL17 was stable at all RIN^e values tested. Furthermore, we investigated the effect of quantity and quality of RNA on the quality of the resultant RNA sequencing (RNA-Seq) data. Thirteen RNA-seq data showed similar duplication and mapping rates (94 to 95%) against the feline genome regardless of RIN^e values. However one low yield sample with a high RIN^e value showed a high duplication rate and it was an outlier on the RNA-seq multidimensional scaling plot. We conclude that the overall yield of RNA was more important than quality of RNA for RNA-seq quality control. These results will guide researchers who wish to perform RNA extractions from mineralised tissues, especially if collecting in a clinical setting with the recognised restraints that this imposes.

Holo-Seq: single-cell sequencing of holo-transcriptome

Current single-cell RNA-seq approaches are hindered by preamplification bias, loss of strand of origin information, and the inability to observe small-RNA and mRNA dual transcriptomes. Here, we introduce a single-cell holo-transcriptome sequencing (Holo-Seq) that overcomes all three hurdles. Holo-Seq has the same quantitative accuracy and uniform coverage with a complete strand of origin information as bulk RNA-seq. Most importantly, Holo-Seq can simultaneously observe small RNAs and mRNAs in a single cell. Furthermore, we acquire small RNA and mRNA dual transcriptomes of 32 human hepatocellular carcinoma single cells, which display the genome-wide super-enhancer activity and hepatic neoplasm kinetics of these cells.

Consensus on Molecular Subtypes of High-Grade Serous Ovarian Carcinoma

Purpose: The majority of ovarian carcinomas are of high-grade serous histology, which is associated with poor prognosis. Surgery and chemotherapy are the mainstay of treatment, and molecular characterization is necessary to lead the way to targeted therapeutic options. To this end, various computational methods for gene expression-based subtyping of high-grade serous ovarian carcinoma (HGSOC) have been proposed, but their overlap and robustness remain unknown.Experimental Design: We assess three major subtype classifiers by meta-analysis of publicly available expression data, and assess statistical criteria of subtype robustness and classifier concordance. We develop a consensus classifier that represents the subtype classifications of tumors based on the consensus of multiple methods, and outputs a confidence score. Using our compendium of expression data, we examine the possibility that a subset of tumors is unclassifiable based on currently proposed subtypes.Results: HGSOC subtyping classifiers exhibit moderate pairwise concordance across our data compendium (58.9%-70.9%; P < 10^-5) and are associated with overall survival in a meta-analysis across datasets (P < 10^-5). Current subtypes do not meet statistical criteria for robustness to reclustering across multiple datasets (prediction strength < 0.6). A new subtype classifier is trained on concordantly classified samples to yield a consensus classification of patient tumors that correlates with patient age, survival, tumor purity, and lymphocyte infiltration.Conclusions: A new consensus ovarian subtype classifier represents the consensus of methods and demonstrates the importance of classification approaches for cancer that do not require all tumors to be assigned to a distinct subtype. Clin Cancer Res; 24(20); 5037-47. ©2018 AACR.

Consensus on Molecular Subtypes of High-Grade Serous Ovarian Carcinoma

Purpose: The majority of ovarian carcinomas are of high-grade serous histology, which is associated with poor prognosis. Surgery and chemotherapy are the mainstay of treatment, and molecular characterization is necessary to lead the way to targeted therapeutic options. To this end, various computational methods for gene expression-based subtyping of high-grade serous ovarian carcinoma (HGSOC) have been proposed, but their overlap and robustness remain unknown.Experimental Design: We assess three major subtype classifiers by meta-analysis of publicly available expression data, and assess statistical criteria of subtype robustness and classifier concordance. We develop a consensus classifier that represents the subtype classifications of tumors based on the consensus of multiple methods, and outputs a confidence score. Using our compendium of expression data, we examine the possibility that a subset of tumors is unclassifiable based on currently proposed subtypes.Results: HGSOC subtyping classifiers exhibit moderate pairwise concordance across our data compendium (58.9%-70.9%; P < 10-5) and are associated with overall survival in a meta-analysis across datasets (P < 10-5). Current subtypes do not meet statistical criteria for robustness to reclustering across multiple datasets (prediction strength < 0.6). A new subtype classifier is trained on concordantly classified samples to yield a consensus classification of patient tumors that correlates with patient age, survival, tumor purity, and lymphocyte infiltration.Conclusions: A new consensus ovarian subtype classifier represents the consensus of methods and demonstrates the importance of classification approaches for cancer that do not require all tumors to be assigned to a distinct subtype. Clin Cancer Res; 24(20); 5037-47. ©2018 AACR.

Neuro-evo-devo in the single cell sequencing era

The nervous system represents the most complex tissue in animals. How this complexity evolved has been a challenging question to address. The explosion in single cell sequencing techniques, the development of new algorithms to cluster single cells into cell types, along with powerful tools for drawing developmental trajectories offer a unique opportunity to compare homologous cell types between species. They further permit the identification of key developmental points and transcription factors that can lead to the evolution of new cell types. At the same time, the ease of use and efficiency of CRISPR genome editing technology allow validation of predicted regulators. This promises exciting developments in the next few years in the field of neuronal evolution and development.

Single-cell transcriptomics of 20 mouse organs creates a Tabula Muris

Here we present a compendium of single-cell transcriptomic data from the model organism Mus musculus that comprises more than 100,000 cells from 20 organs and tissues. These data represent a new resource for cell biology, reveal gene expression in poorly characterized cell populations and enable the direct and controlled comparison of gene expression in cell types that are shared between tissues, such as T lymphocytes and endothelial cells from different anatomical locations. Two distinct technical approaches were used for most organs: one approach, microfluidic droplet-based 3'-end counting, enabled the survey of thousands of cells at relatively low coverage, whereas the other, full-length transcript analysis based on fluorescence-activated cell sorting, enabled the characterization of cell types with high sensitivity and coverage. The cumulative data provide the foundation for an atlas of transcriptomic cell biology.

Missing data and technical variability in single-cell RNA-sequencing experiments

Until recently, high-throughput gene expression technology, such as RNA-Sequencing (RNA-seq) required hundreds of thousands of cells to produce reliable measurements. Recent technical advances permit genome-wide gene expression measurement at the single-cell level. Single-cell RNA-Seq (scRNA-seq) is the most widely used and numerous publications are based on data produced with this technology. However, RNA-seq and scRNA-seq data are markedly different. In particular, unlike RNA-seq, the majority of reported expression levels in scRNA-seq are zeros, which could be either biologically-driven, genes not expressing RNA at the time of measurement, or technically-driven, genes expressing RNA, but not at a sufficient level to be detected by sequencing technology. Another difference is that the proportion of genes reporting the expression level to be zero varies substantially across single cells compared to RNA-seq samples. However, it remains unclear to what extent this cell-to-cell variation is being driven by technical rather than biological variation. Furthermore, while systematic errors, including batch effects, have been widely reported as a major challenge in high-throughput technologies, these issues have received minimal attention in published studies based on scRNA-seq technology. Here, we use an assessment experiment to examine data from published studies and demonstrate that systematic errors can explain a substantial percentage of observed cell-to-cell expression variability. Specifically, we present evidence that some of these reported zeros are driven by technical variation by demonstrating that scRNA-seq produces more zeros than expected and that this bias is greater for lower expressed genes. In addition, this missing data problem is exacerbated by the fact that this technical variation varies cell-to-cell. Then, we show how this technical cell-to-cell variability can be confused with novel biological results. Finally, we demonstrate and discuss how batch-effects and confounded experiments can intensify the problem.

A comparative analysis of library prep approaches for sequencing low input translatome samples

BACKGROUND

Cell type-specific ribosome-pulldown has become an increasingly popular method for analysis of gene expression. It allows for expression analysis from intact tissues and monitoring of protein synthesis in vivo. However, while its utility has been assessed, technical aspects related to sequencing of these samples, often starting with a smaller amount of RNA, have not been reported. In this study, we evaluated the performance of five library prep protocols for ribosome-associated mRNAs when only 250 pg-4 ng of total RNA are used.

RESULTS

We obtained total and RiboTag-IP RNA, in three biological replicates. We compared 5 methods of library preparation for Illumina Next Generation sequencing: NuGEN Ovation RNA-Seq system V2 Kit, TaKaRa SMARTer Stranded Total RNA-Seq Kit, TaKaRa SMART-Seq v4 Ultra Low Input RNA Kit, Illumina TruSeq RNA Library Prep Kit v2 and NEBNext® Ultra™ Directional RNA Library Prep Kit using slightly modified protocols each with 4 ng of total RNA. An additional set of samples was processed using the TruSeq kit with 70 ng, as a 'gold standard' control and the SMART-Seq v4 with 250 pg of total RNA. TruSeq-processed samples had the best metrics overall, with similar results for the 4 ng and 70 ng samples. The results of the SMART-Seq v4 processed samples were similar to TruSeq (Spearman correlation > 0.8) despite using lower amount of input RNA. All RiboTag-IP samples had an increase in the intronic reads compared with the corresponding whole tissue, suggesting that the IP captures some immature mRNAs. The SMARTer-processed samples had a higher representation of ribosomal and non-coding RNAs leading to lower representation of protein coding mRNA. The enrichment or depletion of IP samples compared to corresponding input RNA was similar across all kits except for SMARTer kit.

CONCLUSION

RiboTag-seq can be performed successfully with as little as 250 pg of total RNA when using the SMART-Seq v4 kit and 4 ng when using the modified protocols of other library preparation kits. The SMART-Seq v4 and TruSeq kits resulted in the highest quality libraries. RiboTag IP RNA contains some immature transcripts.

Comparative transcriptomic analyses and single-cell RNA sequencing of the freshwater planarian Schmidtea mediterranea identify major cell types and pathway conservation

BACKGROUND

In the Lophotrochozoa/Spiralia superphylum, few organisms have as high a capacity for rapid testing of gene function and single-cell transcriptomics as the freshwater planaria. The species Schmidtea mediterranea in particular has become a powerful model to use in studying adult stem cell biology and mechanisms of regeneration. Despite this, systematic attempts to define gene complements and their annotations are lacking, restricting comparative analyses that detail the conservation of biochemical pathways and identify lineage-specific innovations.

RESULTS

In this study we compare several transcriptomes and define a robust set of 35,232 transcripts. From this, we perform systematic functional annotations and undertake a genome-scale metabolic reconstruction for S. mediterranea. Cross-species comparisons of gene content identify conserved, lineage-specific, and expanded gene families, which may contribute to the regenerative properties of planarians. In particular, we find that the TRAF gene family has been greatly expanded in planarians. We further provide a single-cell RNA sequencing analysis of 2000 cells, revealing both known and novel cell types defined by unique signatures of gene expression. Among these are a novel mesenchymal cell population as well as a cell type involved in eye regeneration. Integration of our metabolic reconstruction further reveals the extent to which given cell types have adapted energy and nucleotide biosynthetic pathways to support their specialized roles.

CONCLUSIONS

In general, S. mediterranea displays a high level of gene and pathway conservation compared with other model systems, rendering it a viable model to study the roles of these pathways in stem cell biology and regeneration.

NeuroSystematics and Periodic System of Neurons: Model vs Reference Species at Single-Cell Resolution

There is more than one way to develop neuronal complexity, and animals frequently use different molecular toolkits to achieve similar functional outcomes (=convergent evolution). Neurons are different not only because they have different functions, but also because neurons and circuits have different genealogies, and perhaps independent origins at the broadest scale from ctenophores and cnidarians to cephalopods and primates. By combining modern phylogenomics, single-neuron sequencing (scRNA-seq), machine learning, single-cell proteomics, and metabolomic across Metazoa, it is possible to reconstruct the evolutionary histories of neurons tracing them to ancestral secretory cells. Comparative data suggest that neurons, and perhaps synapses, evolved at least 2-3 times (in ctenophore, cnidarian and bilateral lineages) during ∼600 million years of animal evolution. There were also several independent events of the nervous system centralization either from a common bilateral/cnidarian ancestor without the bona fide neurons or from the urbilaterian with diffuse, nerve-net type nervous system. From the evolutionary standpoint, (i) a neuron should be viewed as a functional rather than a genetic character, and (ii) any given neural system might be chimeric and composed of different cell lineages with distinct origins and evolutionary histories. The identification of distant neural homologies or examples of convergent evolution among 34 phyla will not only allow the reconstruction of neural systems' evolution but together with single-cell "omic" approaches the proposed synthesis would lead to the "Periodic System of Neurons" with predictive power for neuronal phenotypes and plasticity. Such a phylogenetic classification framework of Neuronal Systematics (NeuroSystematics) might be a conceptual analog of the Periodic System of Chemical Elements. scRNA-seq profiling of all neurons in an entire brain or Brain-seq is now fully achievable in many nontraditional reference species across the entire animal kingdom. Arguably, marine animals are the most suitable for the proposed tasks because the world oceans represent the greatest taxonomic and body-plan diversity.